1

Compiler DesignType Checking

Santanu Chattopadhyay

Electronics and Electrical Communication Engineering

2

What is a Type Checking Static vs. Dynamic Checking Type Expressions Type Equivalence Type Conversion Phases of a Compiler Conclusion

What is Type Checking• One of the most important semantic aspects of compilation

• Allows the programmer to limit what types may be used incertain circumstances

• Assigns types to values

• Determines whether these values are used in an appropriatemanner

• Simplest situation: check types of objects and report a type-error in case of a violation

• More complex: incorrect types may be corrected (typecoercing)

3

Static vs. Dynamic Checking• Static Checking

– Type checking done at compile time

– Properties can be verified before program run

– Can catch many common errors

– Desirable when faster execution is important

• Dynamic Checking

– Performed during program execution

– Permits programmers to be less concerned with types

– Mandatory in some situations, such as, array bounds check

– More robust and clearer code

4

Type Expressions

• Used to represent types of language constructs

• A type expression can be – Basic type: integer, real, char, Boolean and other atomic types that do

not have internal structure. A special type, type-error is used to indicate type violations

– Type name

– Type constructor applied to a list of type expressions

5

Type Expressions• Arrays are specified as array(I,T), where T is a type and I is an integer or a

range of integers. For example, C declaration “int a[100]” identifies type of a to be array(100, integer)

• If T1 and T2 are type expressions, T1 × T2 represents “anonymous records”. For example, an argument list passed to a function with first argument integer and second real, has type integer × real

6

Type Expressions• Named records are products with named elements. For a record structure with

two named fields – length (an integer) and word (of type array(10, char)), the record is of type

record((length × integer) × (word × array(10, character)))

• If T is a type expression, pointer(T) is also a type expression, representing objects that are pointers to objects of type T

• Function maps a collection of types to another, represented by D R, where D is the domain and R is the range of the function.

7

Type Expressions

• Type expression “integer × integer character” represents a function that takes two integers as arguments and returns a character value

• Type expression “integer (real character)” represents a function that takes an integer as an argument and returns another function which maps a real number to a character

8

Type Systems

• Type system of a language is a collection of rules depicting the type expression assignments to program objects

• Usually done with syntax directed definition

• ‘Type checker’ is an implementation of a type system

9

Strongly Typed Language

• Compiler can verify that the program will execute without any type errors

• All checks are made statically

• Also called a sound type system

• Completely eliminates necessity of dynamic type checking

• Most programming languages are weakly typed

• Strongly typed languages put lot of restrictions

• There are cases in which a type error can be caught dynamically only

• Many languages also allow the user ti override the system

10

Type Checking of Expressions

• Use synthesized attribute ‘type’ for the nonterminal E representing an expression

11

Type Checking of Statements• Statements normally do not have any value, hence of type

void

• For propagating type error occurring in some statement nested deep inside a block, a set of rules needed

12

Type Checking of Functions

• A function call is equivalent to the application of one expression to another

13

Type Equivalence

• It is often needed to check whether two type expressions ‘s’ and ‘t’ are same or not

• Can be answered by deciding equivalence between the two types

• Two categories of equivalence

– Name equivalence

– Structural equivalence

14